Cade Povich was strong in the beginning for the Baltimore Orioles tonight. Despite a 30 minute rain delay to start the game, Povich mowed runners down. Povich’s line: 4.2 IP, 8 H, 5 R, 3 BB, 9 K.
Ryan O’Hearn scored on an error in the wake of Cedric Mullins’ bunt base hit in the second. The Orioles would tack on a run on an RBI-groundout by Maverick Handley later in the inning, and Mullins himself would score on an RBI-double by Jackson Holliday. And the Orioles led 3-0 early.
However St. Louis would get a two-RBI double by Walker in the fourth to cut the lead to 3-2. In the home half of the inning however the Birds would tack on an extra run. Dylan Carlson’s RBI-single would run the lead to 4-2. However back-to-back doubles in the fifth would cut the lead back to one for the Orioles. Donovan’s two-run homer later in the inning gave St. Louis a 5-4 lead, and it would extend to 6-4 in the sixth with a solo homer by Herrera.
I noticed something in this game, that being that more than once St. Louis split defenders on hits, and that the ball would go all the way to the wall. And it’s something I’ve noticed a lot thus far this season. If it happens once or twice you chalk it up to luck or sleight of hand. But…it’s happened a lot.
The Orioles’ use of analytics is well-documented. Part of that is probability. So they position their defense in such a way that makes the probability higher that the opponent will hit the ball to someone’s glove. That seems like a winning formula.
But the computer can’t attest to the fact that opponents read those same stats and probability. Given that they know the Orioles are relying on only these analytics, they’re reading their own stats and tweaking their hitting just enough to change where they’re placing the ball. In short, analytics are an exact science. They have to be. But sports can’t ever be an exact science by their very nature. And we should never treat them as such.
Games are meant to be won on the field between the lines – not by a computer formula. That’s why opposing teams are able to consistently place the ball precisely where the Orioles aren’t standing. They’re using the Orioles’ use of the computer to outsmart the computer, knowing that the computer isn’t capable of adjusting on the Orioles’ behalf.
The same is true with pitching. How often do teams get caught looking on a full count? They know that the Orioles have an aversion to giving in. That being throwing a fastball on 3-2. So they’re confident that the pitch is going to be out of the strike zone. So they aren’t swinging. And thus they rarely get caught looking.
Analytics are a good thing, to be clear. I’m not a risk-taker in games. Play it safe, always. But this game is still played by men. By humans. It’s the same as relying on AI for anything. You might get exactly what you think you’re getting or are designed to get. But AI can’t predict any variation in the program. Which means it’s easily outdone.
Birdland Crush 